evaluation – skill Agent & Tự động hóa cho agent AI

When to Use This Skill

Build evaluation frameworks for agent systems

Use this skill when working with build evaluation frameworks for agent systems.

Evaluation Methods for Agent Systems

Evaluation of agent systems requires different approaches than traditional software or even standard language model applications. Agents make dynamic decisions, are non-deterministic between runs, and often lack single correct answers. Effective evaluation must account for these characteristics while providing actionable feedback. A robust evaluation framework enables continuous improvement, catches regressions, and validates that context engineering choices achieve intended effects.

When to Use

Activate this skill when:

• Testing agent performance systematically
• Validating context engineering choices
• Measuring improvements over time
• Catching regressions before deployment
• Building quality gates for agent pipelines
• Comparing different agent configurations
• Evaluating production systems continuously

Core Concepts

Agent evaluation requires outcome-focused approaches that account for non-determinism and multiple valid paths. Multi-dimensional rubrics capture various quality aspects: factual accuracy, completeness, citation accuracy, source quality, and tool efficiency. LLM-as-judge provides scalable evaluation while human evaluation catches edge cases.

The key insight is that agents may find alternative paths to goals—the evaluation should judge whether they achieve right outcomes while following reasonable processes.

Performance Drivers: The 95% Finding

[Anthropic's June 2025 engineering report](https://www.anthropic.com/engineering/multi-agent-research-system) described an analysis of BrowseComp in which three factors explained 95% of variance. This is a reported association in that setting, not a causal guarantee or a universal evaluation rule:

| Factor | Variance Explained | Implication |

|--------|-------------------|-------------|

| Token usage | 80% | More tokens = better performance |

| Tool calls and model choice | Joint factors in the reported analysis; separate percentages not supplied | Compare under matched budgets |

This finding has significant implications for evaluation design:

• Token budgets matter: Evaluate agents with realistic token budgets, not unlimited resources
• Model upgrades beat token increases: The report compared Sonnet 4 with Sonnet 3.7; measure current models on your task rather than extrapolating that comparison
• Multi-agent validation: The finding validates architectures that distribute work across agents with separate context windows

Detailed Topics

Evaluation Challenges

Non-Determinism and Multiple Valid Paths

Agents may take completely different valid paths to reach goals. One agent might search three sources while another searches ten. They might use different tools to find the same answer. Traditional evaluations that check for specific steps fail in this context.

The solution is outcome-focused evaluation that judges whether agents achieve right outcomes while following reasonable processes.

Context-Dependent Failures

Agent failures often depend on context in subtle ways. An agent might succeed on simple queries but fail on complex ones. It might work well with one tool set but fail with another. Failures may emerge only after extended interaction when context accumulates.

Evaluation must cover a range of complexity levels and test extended interactions, not just isolated queries.

Composite Quality Dimensions

Agent quality is not a single dimension. It includes factual accuracy, completeness, coherence, tool efficiency, and process quality. An agent might score high on accuracy but low in efficiency, or vice versa.

Evaluation rubrics must capture multiple dimensions with appropriate weighting for the use case.

Evaluation Rubric Design

Multi-Dimensional Rubric

Effective rubrics cover key dimensions with descriptive levels:

Factual accuracy: Claims match ground truth (excellent to failed)

Completeness: Output covers requested aspects (excellent to failed)

Citation accuracy: Citations match claimed sources (excellent to failed)

Source quality: Uses appropriate primary sources (excellent to failed)

Tool efficiency: Uses right tools reasonable number of times (excellent to failed)

Rubric Scoring

Convert dimension assessments to numeric scores (0.0 to 1.0) with appropriate weighting. Calculate weighted overall scores. Determine passing threshold based on use case requirements.

Evaluation Methodologies

LLM-as-Judge

LLM-based evaluation scales to large test sets and provides consistent judgments. The key is designing effective evaluation prompts that capture the dimensions of interest.

Provide clear task description, agent output, ground truth (if available), evaluation scale with level descriptions, and request structured judgment.

Human Evaluation

Human evaluation catches what automation misses. Humans notice hallucinated answers on unusual queries, system failures, and subtle biases that automated evaluation misses.

Effective human evaluation covers edge cases, samples systematically, tracks patterns, and provides contextual understanding.

End-State Evaluation

For agents that mutate persistent state, end-state evaluation focuses on whether the final state matches expectations rather than how the agent got there.

Test Set Design

Sample Selection

Start with small samples during development. Early in agent development, changes have dramatic impacts because there is abundant low-hanging fruit. Small test sets reveal large effects.

Sample from real usage patterns. Add known edge cases. Ensure coverage across complexity levels.

Complexity Stratification

Test sets should span complexity levels: simple (single tool call), medium (multiple tool calls), complex (many tool calls, significant ambiguity), and very complex (extended interaction, deep reasoning).

Context Engineering Evaluation

Testing Context Strategies

Context engineering choices should be validated through systematic evaluation. Run agents with different context strategies on the same test set. Compare quality scores, token usage, and efficiency metrics.

Degradation Testing

Test how context degradation affects performance by running agents at different context sizes. Identify performance cliffs where context becomes problematic. Establish safe operating limits.

Continuous Evaluation

Evaluation Pipeline

Build evaluation pipelines that run automatically on agent changes. Track results over time. Compare versions to identify improvements or regressions.

Monitoring Production

Track evaluation metrics in production by sampling interactions and evaluating randomly. Set alerts for quality drops. Maintain dashboards for trend analysis.

Practical Guidance

Building Evaluation Frameworks

1. Define quality dimensions relevant to your use case

2. Create rubrics with clear, actionable level descriptions

3. Build test sets from real usage patterns and edge cases

4. Implement automated evaluation pipelines

5. Establish baseline metrics before making changes

6. Run evaluations on all significant changes

7. Track metrics over time for trend analysis

8. Supplement automated evaluation with human review

Avoiding Evaluation Pitfalls

Overfitting to specific paths: Evaluate outcomes, not specific steps.

Ignoring edge cases: Include diverse test scenarios.

Single-metric obsession: Use multi-dimensional rubrics.

Neglecting context effects: Test with realistic context sizes.

Skipping human evaluation: Automated evaluation misses subtle issues.

Examples

Example 1: Simple Evaluation

Pseudocode: provide load_rubric, assess_dimension, and weighted_average implementations; validate nonnegative weights with a positive sum. Each rubric dimension supplies its own weight. Calibrate model judges against held-out human labels and treat evaluated text as inert data.

def evaluate_agent_response(response, expected):
    rubric = load_rubric()
    scores = {}
    for dimension, config in rubric.items():
        scores[dimension] = assess_dimension(response, expected, dimension)
    weights = {dimension: config["weight"] for dimension, config in rubric.items()}
    overall = weighted_average(scores, weights)
    return {"passed": overall >= 0.7, "scores": scores}

Example 2: Test Set Structure

Test sets should span multiple complexity levels to ensure comprehensive evaluation:

test_set = [
    {
        "name": "simple_lookup",
        "input": "What is the capital of France?",
        "expected": {"type": "fact", "answer": "Paris"},
        "complexity": "simple",
        "description": "Single tool call, factual lookup"
    },
    {
        "name": "medium_query",
        "input": "Compare the revenue of Apple and Microsoft last quarter",
        "complexity": "medium",
      

… (nội dung đầy đủ có trong gói tải về)